Anthropic Settles Copyright Dispute with Authors
Background to the Dispute: Allegations of Stealing Millions of Books
It all began when a group of American authors and publishers filed a class-action lawsuit against Anthropic. They claimed that the company had downloaded and used up to seven million books from pirate websites to train its AI models. According to them, this action infringed the copyrights of a large number of works, which could lead to enormous financial penalties—ranging from $750 to $150,000 for each infringement. Judge William Alsup, who presided over the case, rejected Anthropic's request to stay the proceedings, forcing the company to face the reality of potential liability amounting to billions of dollars. This dispute was the first certified class-action lawsuit against an AI company in the field of copyright, making it a precedent for similar cases against companies such as OpenAI, Microsoft, and Meta.
Key Development: Court Ruling on Fair Use
Before the settlement itself, the court considered whether the use of books to train AI fell under the concept of "fair use," a U.S. legal doctrine allowing limited use of protected materials without permission. The court ruled that Anthropic's use of books as training data constituted fair use, which was a victory for the company. However, that was not enough to bring the dispute to a complete end—the plaintiffs continued their fight, emphasizing that pirated downloading and copying exceeded the bounds of legality. The ruling came after intensive hearings that discussed technical details, such as how AI models process data from books without directly copying their content.
Settlement on the Horizon: Details and Timeline
Now to the heart of the matter—Anthropic and the plaintiffs informed the federal court in California that they had reached an agreement to settle the entire class-action lawsuit. The terms are not yet public, but they are expected to be disclosed in early September 2025 following court approval. Judge Alsup ordered the parties to file a motion for preliminary approval by September 5, with a hearing scheduled for September 8, 2025. The agreement has been described as "historic" by organizations such as the Authors Guild and the Association of American Publishers, which hope it will push the AI industry toward more ethical practices and fair compensation for creators. Although the settlement does not establish a formal legal precedent, it may influence strategies in future disputes.
Other Disputes and Reactions
Anthropic still faces other lawsuits, including those filed by music publishers and the Reddit platform, which accuse the company of similar unauthorized use of protected content. According to information available from websites such as Publishers Weekly and Fortune, this settlement could be the first domino to fall in a series of AI copyright battles. Experts note that although the financial details are unknown, the distribution of compensation among authors and publishers will be crucial. Industry reactions are mixed—while creators are celebrating a step forward, AI companies see it as a signal to take a more cautious approach to data.
Source: www.theinformation.com
